Bringing Home the Bacon: An Empirical Analysis of the Extent and Effects of PorkBarrelling in Australian Politics*
Andrew Leigh
Social Policy Evaluation, Analysis and Research Centre
Research School of Social Sciences
Australian National University
[email protected]
http://econrsss.anu.edu.au/~aleigh/
This draft: May 2006
Abstract
Which electorates receive targeted funding, and does targeted funding swing votes? To
answer these questions, I focus on four discretionary programs introduced by the
Australian federal government during the 2001-04 election cycle: Roads to Recovery,
Stronger Families and Communities, Sustainable Regions, and Regional Partnerships.
Together, these programs totalled well over $1 billion dollars. Controlling for relevant
demographic characteristics of the electorate, I find that those electorates held by the
governing Liberal-National Coalition (particularly those held by the National Party)
received a larger share of discretionary funding, and a larger number of program grants.
Among government seats, I find no evidence that funding was directed towards those that
were more marginal. More discretionary funding – particularly from the Roads to
Recovery program – was associated with a larger swing towards the Coalition in the 2004
election, indicating that targeted funding was rewarded at the ballot box.
JEL Codes: D72, R58
Keywords: elections, local expenditure, voting, targeted funding, pork barrelling
*
I am grateful to Mark Davis, Mathew Jose, Marie Liau, and Matthew Linden for providing me with some
of the data used in this paper, and to Sinclair Davidson, Tim Fry, Murray Goot, Elena Varganova and Justin
Wolfers for valuable comments on earlier drafts. I have also benefited from feedback received at a
Democratic Audit of Australia workshop, a conference on John Howard’s Decade, and a seminar at RMIT
University.
1. Introduction
Pork barrel politics – the practice of targeting expenditure to particular districts based on
political considerations – has been in existence for at least two centuries.1 In the United
States, where the term was originally coined, over 15,000 projects per year are
“earmarked” for particular districts (Flake 2006). Variously sponsored by representatives
on both sides of the politics, such projects are frequently added onto budget
appropriations to accommodate constituents, campaign donors, or potential supporters.
While pork-barrelling has been extensively studied in the US, less research has been
conducted on the phenomenon among parliamentary democracies, in which political
parties typically exert more control over their legislators. When decision-making over
local expenditure is more highly centralized, resources may be allocated differently than
in a decentralized system.
Here, I focus on four multi-million dollar Australian programs that were allocated on a
discretionary basis across federal electorates. As a parliamentary democracy with
compulsory voting, Australia provides a useful testing-ground for theories about the
partisan allocation of discretionary funding – both between government and opposition,
and within the governing Coalition. It also provides an opportunity to explore the effect
of additional spending on voting patterns in the subsequent election.
To preview my findings, I observe a strong partisan component to the expenditure
decisions, with more generous funding and more program grants allocated to electorates
held by the party in power. This result is robust to controlling for demographic
characteristics of the electorate that might have affected the allocation of funding.
Estimating the effect of this expenditure on voting, I find targeted funding – particularly
1
Gordon (1993) argues that the construction of the Egyptian pyramids was a form of pork-barrelling,
designed to keep peasants from rebelling. A more settled example is the US Bonus Bill 1817, a highway
project introduced by then Congressman John Calhoun, but ultimately vetoed by President Madison. The
Oxford English Dictionary dates the first use of the term ‘pork barrel’ to an article written in the
Westminster Gazette in 1909.
2
roads funding – had a positive and statistically significant impact on the swing received
by the governing Coalition in the 2004 election.
The remainder of this paper is organised as follows. Section 2 briefly reviews the relevant
literature. Section 3 outlines the details of the four funding programs and relevant
electoral data. Section 4 estimates the extent to which funding decisions appear to have
been skewed by political considerations. Section 5 estimates the effect of the funding on
the results in the following election, and the final section concludes.
2. Research on Pork-Barrelling
The ability of governments to apportion local-level expenditure for partisan purposes has
long been of interest to political scientists and economists. One of the key questions in
this literature is whether politicians allocate resources primarily towards swing seats or
core supporters. While Cox and McCubbins (1986) posited a model in which politicians
are risk-averse, and therefore channel resources more generously towards their core
supporters, Dixit and Londregan (1996) argued that in certain circumstances, politicians
may prefer to spend money on swing voters.
Several studies have sought to determine whether a greater share of spending is directed
towards core supporters or swing voters. Programs that seem to be more targeted towards
core supporters include Canadian regional development grants (Milligan and Smart 2005)
and New Deal funding in the US during the 1930s (Lindstädt 2005). By contrast, those
that appear to be more targeted towards swing voters include Canadian job training grants
(Crampton 2004), French road spending (Cadot et al 2002), allocations to Indian states
(Dasgupta et al 2002), and Swedish environmental spending (Dahlberg and Johansson
2002).2
2
Other studies have looked at the manner in which committee memberships affect the distribution of porkbarrelling in the US (eg. Alvarez and Saving 1997a; Stein and Bickers 1995).
3
In the Australian context, two studies of a $60 million sports grants program in the early1990s concluded that the spending was directed in a partisan fashion, primarily towards
swing voters (Gaunt 1999; Denemark 2000). Similarly, an analysis of federal programs
for the unemployed (Andrews, Fry and Jakee 2005) found that the program was skewed
towards government-held marginal electorates.
In general, one should expect that in systems with weaker political parties and stronger
individual members, pork-barrelling will tend to be directed towards core supporters. By
contrast, when parties are stronger and individual members weaker, discretionary funding
is more likely to be targeted towards electorates with a smaller vote margin.3 As
Denemark (2000, 898) has noted:
Unlike the sharp dichotomy between the interests of the individual and the party
that fuels constituency activities in America, parliamentary parties in government
confront a collective electoral imperative to assure the victory of their most
vulnerable party colleagues in marginal seats. In short, the parliamentary gap
between individual and collective interests is “virtually nonexistent” (Cain,
Ferejohn, and Fiorina 1984, 111).
With some exceptions (eg. Milligan and Smart 2005), the empirical literature across
countries has tended to support this theoretical prediction.
How much does pork-barrelling matter at the ballot box? While some studies have
observed little or no relationship between local expenditure and voteshare (Feldman and
Jondrow 1984; Stein and Bickers 1994), others have found that more spending raises the
voteshare of the incumbent (Alvarez and Saving 1997b; Levitt and Snyder 1997). Levitt
and Snyder (1997) suggested that if spending levels are higher when the incumbent is
weak, the coefficient on expenditure might be biased downwards, and instrument for
3
An additional factor in Australia is the existence of compulsory voting, which substantially reduces the
risk that the governing party’s core supporters will refuse to vote.
4
local spending using spending in the same state but outside the district. I discuss this
possible bias in section 5.
3. The Programs and Electoral Data
In selecting programs, my focus is on Australian federal government programs that meet
the following criteria:
(a) they are regional in nature, and allocated in a manner that allowed for some
discretion by politicians;
(b) grants were largely announced or delivered between the 10 November 2001 and
16 October 2004 federal elections; and
(c) statistics on the size of the funding are available on by electorate.
Four programs meet these three criteria: Roads to Recovery, Stronger Families and
Communities, Sustainable Regions, and Regional Partnerships. In sections 3.1 to 3.4, I
discuss each of the programs, focusing on the aims of the programs, the dates when the
funding was delivered, the funding criteria, the source of the funding data, and any
publicly available cost-benefit analyses. Cost-benefit studies are relevant since a narrow
definition of pork-barrelling requires not only that the allocation be based primarily on
political considerations, but also that the program is “economically inefficient”
(Lancaster and Patterson 1990). Section 3.5 outlines the electoral and demographic
variables.
3.1. Roads to Recovery
The Roads to Recovery program was announced in November 2000, and commenced in
January 2001. It provides funding to local councils to undertake “the construction,
upgrade or maintenance of roads”. Overall, 46 percent of funding was devoted to
reconstruction, rehabilitation or widening of existing roads, 26 percent to sealing or
resealing, 7 percent to bridges and tunnels, and 6 percent to the construction of a new
5
road. The remainder of the funding was devoted to smaller programs, including signage,
street lighting and bicycle paths (DOTARS/ALGA 2003).
To prevent federal funding merely being substituted for state funding, a condition of
receipt is that local councils are required to maintain their roads expenditure at the same
level as in the 1998-99 to 2000-01 financial years. Another condition requires that each
of the projects be signposted at both ends with a “Roads to Recovery” sign,
acknowledging the financial assistance of the federal government (with each pair of signs
costing $550, the total cost of this advertising across more than 8000 projects nationwide
exceeds $4 million). The allocation mechanism is somewhat opaque, but according to the
program’s annual report (DOTARS 2004), funding is allocated across states and
territories according to “historical precedents, length of local roads and population”. The
report does not make clear precisely how funding is allocated across local areas.
In this analysis, I focus on $1.1 billion of funding allocated between January 2001 and
June 2005 ($1.2 billion was allocated, but not all projects could be matched to a federal
electorate). For the most part, this funding was announced and delivered during the 200104 election cycle. Figures were tabulated according to local councils (DOTARS 2004),
and were then matched to federal electorates by Davis (2005).4
In 2004, the federal government commissioned a cost-benefit analysis of the programs.
This covered 80 projects, slightly less than 1 percent of the total number of Roads to
Recovery projects funded to that point. (It was not clear on what basis these projects were
selected for inclusion in the analysis.) Based on data provided by councils, the analysis
concluded that the average benefit-cost ratio of Roads to Recovery projects was 1.8.
However, the distribution was highly skewed. While some projects had benefit-cost ratios
as high as 16, most were much lower. Fifty-five percent of projects had a benefit-cost
4
In an email to me, Davis stated that “The Roads to Recovery figures were reported on a Local
Government Area basis because the money goes straight to local councils. I converted those figures firstly
to statistical local areas and then from SLAs to federal electorates, allocating the LGA spending amounts
according to population distributions.” (email communication, 10 March 2005). Davis did not tabulate
figures for the ACT and the NT, each of which were allocated $20 million. In the absence of further
detailed breakdowns, I assume that the seats of Canberra (ACT), Fraser (ACT), Lingiari (NT) and Solomon
(NT) each received $10 million.
6
ratio below 1, so for the median Roads to Recovery program, the costs outweighed the
benefits. Only 20 percent of programs had a benefit-cost ratio exceeding 2. By contrast,
the National Black Spot program, which operated over this period, refused to consider
applications for any programs whose benefit-cost ratio was lower than 2 (DOTARS
2001a, Part 2.1)
3.2. Stronger Families and Communities
The Stronger Families and Communities Strategy commenced in April 2000. The
strategy comprised four initiatives: (i) Communities for Children; (ii) Invest to Grow–
Established and Developing Programs; (iii) Local Answers; and (iv) Choice and
Flexibility in Child Care. Here, I focus on Volunteer Small Equipment Grants delivered
under the “Local Answers” program in the financial year 2004-05. Although the 2004
federal election was held early in the financial year, these grants were typically
announced or delivered prior to the election.
The aim of the Local Answers program was to supports projects that: “build effective
parenting and relationships skills; build opportunities and skills for economic self
reliance in families and communities; strengthen support to families and communities by
delivering better services and addressing unmet needs through the building of
partnerships between local services; assist young parents in particular to further their
education or access to training and other services where they are seeking to make the
transition to employment; assist members of the community to get involved in
community life through local volunteering or mentoring of young people or training to
build community leadership and initiative.” (FACS 2005a). So far as I am aware, no costbenefit analysis of the Local Answers program was conducted.5
5
The 2000-04 evaluation of the Stronger Families and Communities strategy was primarily qualitative in
nature. A summary of the evaluation stated: “The key evaluation questions for Local Answers are: to what
extent were aims and outcomes achieved? what were the success factors? what were the unintended project
outcomes and how well were they addressed?” (FACS 2005b).
7
Volunteer Small Equipment Grants were grants to “encourage and support volunteers by
enabling local community organisations to purchase small equipment items to make the
work of their volunteers easier, safer and more enjoyable” (FACS 2005a). The funding
data, tabulated by electorate, was provided by the Department of Family and Community
Services in response to a question on notice by Senator Chris Evans.6 It covers 1206
grants, ranging in value from $59 to $4545.7 For example, grants included $2200 to the
Apex Club of Bairnsdale, $1833 to the Munglinup Pony Club, $2932 to the Ryde
Regional Radio Cooperative, and $1204 to the Blackburn Community Church of Christ.
Nationally, the grants amounted to $2.5 million; only a small fraction of the total
expenditure on the Stronger Families and Communities strategy. These grant-level
figures are then collapsed to electorate level.
3.3 Sustainable Regions
The Sustainable Regions program was announced on 29 August 2001, and was aimed to
assist regions facing major economic, social or environmental change. Initially, eight
regions were selected, and each was allocated a maximum amount, to be spent between
2001 and 2006. The regions selected and the maximum amount allocated were: the
Atherton Tablelands, Queensland ($18 million); Wide Bay Burnett, Queensland ($8
million); Campbelltown-Camden, NSW ($12 million); Far North East NSW ($12
million); Cradle Coast, Tasmania ($12 million); Gippsland, Victoria ($12 million);
Kimberley, WA ($12 million); and Playford/Salisbury, SA ($12 million). In 2004, two
additional regions were selected, to receive funding over the period 2004-08. These were
Northern Rivers and North Coast NSW ($12 million); and Western Qld and Western
NSW ($21 million). These regions in turn allocate funding to specific programs in their
local area, focusing on particular regional priorities. Common priority areas include
creating jobs, attracting new industries, attracting more young people to the area, and
boosting tourism.
6
Senate Community Affairs Legislation Committee, Answers to Estimates Questions on Notice, Family
and Community Services Portfolio, 2004-05 Supplementary Budget Estimates - December 2004, Stronger
Families and Communities Strategy Question 31(k), Attachment A (provided in February 2005).
7
Although FACS (2005) stated that the grants were to be for amounts “up to $3000”, one-quarter of the
grants in the electoral breakdown were for amounts larger than $3000.
8
According to the government department responsible, a variety of considerations were
taken in account in selecting the regions:
“Regions (including urban fringe areas as well as those outside capital cities) were
identified against criteria that included remoteness as well as important socioeconomic and demographic indicators, such as levels of unemployment, family
income and structural change indices, amongst others. Importantly, each of the
eight regions selected initially demonstrated a strong degree of initiative, selfreliance and commitment to community action.” (quoted in SFPAC 2005, para
9.7)
Ultimately, however, final decisions on the program were the responsibility of the
Minister. As the guidelines stated:
“The Sustainable Regions Programme is a discretionary grants programme. The
funding of projects is at the discretion of the Minister for Transport and Regional
Services. Therefore, meeting the assessment criteria and addressing one or more
regional priorities does not guarantee funding.” (quoted in SFPAC 2005, para
9.12)
A Senate Report on the program concluded that “The Committee has been unable to
discover the process by which sustainable regions were selected by the minister.”
(SFPAC 2005, para 9.7). So far as I am aware, no cost-benefit analysis of the Sustainable
Regions program has been conducted.
In this paper, I analyse electorate-level tabulations provided by the Department of
Transport and Regional Services to the Senate Finance and Public Administration
Committee. These tabulations cover individual programs funded under the Sustainable
Regions program. For example, projects include $550,000 given to the North Coast
Plywood Products Pty Ltd (NSW) to build new age plywood preservative treatment
9
facilities; $950,000 given to Burra Foods Pty Ltd (Victoria) to build a cheese Plant;
$15,000 to the River of Gold Slate Mine Pty Ltd (Queensland) to develop a marketing
plan, and $1.1 million to the City of Salisbury (South Australia) for the Wyatt Road
Redevelopment Project. Projects were approved between April 2002 and December 2004.
The data cover 192 projects, totalling $136 million. Of this, $64 million (181 projects)
was spent on projects within a single federal electorate, while the remainder went to
projects that were split across multiple federal electorates. In this analysis, I use only the
$64 million that can be allocated to a single federal electorate, though the results shown
below are similar if the split funding is used instead.
3. 4 Regional Partnerships
The Regional Partnerships program, which commenced on 1 July 2003, is aimed to give
effect to the principles set out in the government’s statement Stronger Regions, A
Stronger Australia (DOTARS 2001b). The statement set out broad goals for regional
assistance, emphasising the desirability of different tiers of government working together,
increasing the economic diversity of regional areas, and communities themselves
identifying the programs they wished to be funded.
The Regional Partnerships program replaced eight precursor programs (the Regional
Solutions, Regional Assistance, Rural Transaction Centres, and Dairy Regional
Assistance programs and four regional structural adjustment programs). Between 1 July
2003 and 31 December 2004, $124 million of expenditure was approved under the
program (SFPAC 2005, para 2.39). The programs are administered by regional bodies
known as Area Consultative Committees (ACCs were first established in 1995; there are
56 in Australia).
Similarly to the Sustainable Regions program, the guidelines for the Regional
Partnerships program explicitly stated that:
10
“Regional Partnerships is a discretionary programme. The funding of projects,
through Regional Partnerships, is at the discretion of the Federal Minister for
Transport and Regional Services or the Federal Minister for Regional Services,
Territories and Local Government, therefore meeting the assessment criteria does
not guarantee funding.” (quoted in SFPAC 2005, para 2.32)
Among the programs funded were $250,000 to develop a complex in Georgetown,
Tasmania to house the Bass and Flinders replica ship; $550,000 for the Slim Dusty
Foundation Ltd in Kempsey (NSW); $1.5 million for dredging work at the mouth of
Tumbi Creek (NSW); and $12.7 million for transitional support to the sugar industry. A
Senate Committee raised questions about the high proportion of applications that were
approved immediately prior to the 2004 federal election. So far as I am aware, no costbenefit analysis of the Regional Partnerships program has been conducted.
In this paper, I analyse electorate-level tabulations provided by the Department of
Transport and Regional Services to the Senate Finance and Public Administration
Committee. These tabulations cover 511 projects, amounting to $111 million of the total
$124 million allocated between July 2003 and December 2004.
3.5 Electoral and Demographic Variables
In both the 2001 and 2004 elections, there were 150 seats in the House of
Representatives. However, as a result of redistributions that occurred in Queensland,
South Australia and Victoria, two seats were abolished (Bonython, SA and Burke, Vic),
and two new seats were created (Bonner, Qld and Gorton, Vic). Further complicating
matters, data on the Roads to Recovery program is tabulated according to 2004
electorates, while data on the Stronger Families and Communities, Sustainable Regions
and Regional Partnerships programs is tabulated according to 2001 electorates.
11
For the purposes of analysing Roads to Recovery expenditure (Section 4), I use the share
of the vote received by the Coalition in 2001.8 For this purpose, it is necessary to impute
voteshare to those seats that did not exist at the 2001 election. This is done using
predictions from election analyst Antony Green, estimated prior to the 2004 election
based on booth-level data from the 2001 election.9 Since the booth-level data suggest that
Labor would have won both seats in 2001, both are classified as Labor seats. The actual
composition of the 150-seat parliament following the 2001 election was 69 Liberal, 13
National, 63 Labor, 3 Independent, 1 Green, and 1 Country Liberal. For the purpose of
this analysis, the Country Liberal Party seat is coded as a National Party seat.
For the purpose of analysing the effect of targeted funding on election outcomes (Section
5), I calculate the “swing” towards or away from the Coalition in each electorate. For this
purpose, it is desirable to measure the swing not by comparing the actual 2001 and 2004
results, but instead by using the booth-level results from 2001 to calculate what the
election result would have been if the 2001 election had been held on the 2004
boundaries, and then comparing this to the actual 2004 election result.10 The variable
Swing is therefore the difference between the 2001 result (on 2004 boundaries) and the
2004 result.11
Demographic variables are included to take into account possible characteristics of the
seat that might be correlated with funding allocation decisions. The variables selected are
the population density (square kilometres per person), and the median family income (in
8
Throughout this paper, references to the Coalition include the Liberal Party, the National Party, and the
Northern Territory Country Liberal Party.
9
Available at www.abc.net.au/elections/federal/2004/. For the seats of Calare, Cunningham, Kennedy and
New England, Antony Green’s two-party preferred estimates are not based on the two major parties. For
these seats, I therefore use the actual 2001 two-party preferred vote instead.
10
For the two seats that were eliminated in 2003, the 2004 vote was calculated as the average of the main
seats into which the former 2001 seat was absorbed. Thus the hypothetical 2004 vote for Bonython was
calculated as 46.2%, being the average of the two-party preferred vote in Makin, Port Adelaide and
Wakefield, while the hypothetical 2004 vote for Burke was calculated as 44.2%, being the average of the
two-party preferred vote for Gorton, Lalor and McEwen.
11
In calculating the swing in this manner, I assume that the redistribution was exogenous with respect to
funding allocations. This is likely to be a reasonable assumption, since the final decision on the new
redistribution is made by the independent Australian Electoral Commission (AEC), giving regard primarily
to population trends and natural boundaries between communities. Although the parties had an opportunity
to make submissions to the AEC, such submissions are much more likely to have focused on the underlying
demographics than on targeted funding decisions.
12
thousands of dollars). Demographic characteristics are from Kopras (2003), based on the
2001 Census and 2001 electoral boundaries.12
Table 1 presents summary statistics. The variables Liberal Party Seat and National Party
Seat denote respectively whether the seat was won by one of the two governing parties in
2001. Complete data for each electorate are presented in the Data Appendix.
Table 1: Summary Statistics
Variable
Roads to Recovery ($m)
Stronger Families ($m)
Stronger Families (# of projects)
Sustainable Regions ($m)
Sustainable Regions (# of projects)
Regional Partnerships ($m)
Regional Partnerships (# of projects)
National Party seat (2001)
Liberal Party seat (2001)
Coalition share of 2PP vote (2001)
Swing (2004 minus 2001)
Population density (km2 per person)
Median weekly family income
($000s)
Obs
150
150
150
150
150
150
150
152
152
152
152
152
Mean
7.845
0.016
7.973
0.428
1.207
0.736
3.353
0.092
0.454
50.763
2.206
0.353
152
0.970
SD
Min
9.712
1.691
0.016
0.000
7.595
0.000
1.811
0.000
5.141
0.000
1.641
0.000
5.012
0.000
0.290
0.000
0.500
0.000
10.862 24.920
5.986 -11.950
1.484
0.000
0.242
0.618
Max
59.642
0.097
49.000
12.928
44.000
12.145
33.000
1.000
1.000
73.930
21.510
10.000
1.792
Note: Between the 2001 and 2004 elections, two seats were abolished and two new seats were created.
Since Roads to Recovery funding is tabulated on 2004 electoral boundaries, and the other three programs
are tabulated on the 2001 electoral boundaries, the summary statistics cover 152 seats. All regression
results, however, cover either 150 seats (if focusing on a single program), or 148 seats (if focusing on
multiple programs).
4. An Empirical Analysis of the Distribution of Funding
4.1 Quantum of funding
To see the relationship between political considerations and funding decisions, Figure 1
charts the percentage of the two-party preferred vote received by the Coalition at the
2001 election against the amount of funding received under each of the programs. All
four programs were more generous to Coalition-held seats (those to the right of the
12
For the two electorates that were created in the 2003 redistribution, the density and income variables take
the mean of the main electorates covering that area in the 2001 election. Thus Bonner is the average for
Bowman and Griffith, while Gorton is the average for Burke, Calwell and Maribyrnong.
13
dashed line) than to those held by the non-government parties (those to the left of the
dashed line). Among Coalition-held seats, it does not appear that more funding was
devoted to marginal seats than safe seats.
Figure 1: Seat Margin and Amount of Targeted Funding Received
0
0
Millions of dollars
5
10
15
Sustainable Regions
Millions of dollars
20
40
60
Roads to Recovery
20
40
60
Coalition 2PP Vote in 2001
80
20
80
Stronger Families and Communities
0
Millions of dollars
0 .02 .04 .06 .08 .1
Millions of dollars
5
10
15
Regional Partnerships
40
60
Coalition 2PP Vote in 2001
20
40
60
Coalition 2PP Vote in 2001
80
20
40
60
Coalition 2PP Vote in 2001
80
To formally test whether the allocation of funding was affected by partisan factors, I
regress the amount of funding assigned to each seat on indicator variables denoting
whether the seat was held by the National Party or the Liberal Party in 2001. Table 2,
Panel A shows the results from this exercise. The largest partisan effects are observed for
Roads to Recovery and Stronger Families and Communities. In the case of Roads to
Recovery, an average of $14.9 million more funding was allocated to each National Party
seat, while $2.9 million more funding was allocated to each Liberal Party seat (both
significant at the 5% level or better). Stronger Families and Communities allocations
were also more generous to National Party Seats (an additional $0.02 million) and to
Liberal Party seats (an additional $0.007 million), a result that is significant at the 1
percent level. More funding was received by National Party seats under the Sustainable
Regions program (an additional $1.7 million) and the Regional Partnerships program (an
14
additional $1.5 million). These two results are statistically significant only at the 10
percent level.
In Panel B, I control for the demographic characteristics of the electorates that might be
associated with the allocation of funding. Specifically, I include a quadratic in population
density for the Roads to Recovery, Sustainable Regions and Regional Partnerships
programs, and a quadratic in income for the Stronger Families and Communities grants.
With these controls, the allocations to the National Party and the Liberal Party remain
statistically significant for the Roads to Recovery program, although the magnitude of the
coefficients falls to +$6.8 million for National Party electorates, and +$2.7 million for
Liberal Party electorates. The population density controls confirm that the program was
indeed more generous towards less densely populated electorates.
For the Stronger Families and Communities program, including income controls has little
impact on the magnitude or statistical significance of the partisan coefficients. The
income controls indicate that slightly more funding was provided to poorer seats.
Controlling for population density, allocations to the Sustainable Regions and Regional
Partnerships do not have a statistically significant partisan bias. The population density
coefficient is statistically significant for the Regional Partnerships program. For the
Sustainable Regions program, the population density coefficients are not statistically
significant.
One point to note about all of the above results is that in all specifications in Panels A and
B, the National Party coefficients are larger in magnitude than the Liberal Party
coefficients. Since the National Party are the junior party in the federal Coalition, this is
somewhat surprising, and suggests that the party’s influence in obtaining targeted funding
was disproportionate to its representation in the government.
15
Table 2: Partisanship and Funding Allocation
Dependent Variable: Total funding per electorate ($m)
Panel A: Without demographic controls
(1)
(2)
R2R
SF&C
National Party
14.904***
0.021***
[4.084]
[0.005]
Liberal Party
2.902**
0.007***
[1.346]
[0.002]
Observations
150
150
R-squared
0.18
0.15
Panel B: With demographic controls
National Party
6.767***
0.015***
[2.152]
[0.005]
Liberal Party
2.699***
0.008***
[0.928]
[0.002]
Income
-0.072**
[0.030]
Income2
0.022*
[0.012]
Population Density
17.704***
[3.320]
2
Population Density
-1.558***
[0.359]
Observations
150
150
R-squared
0.70
0.26
(3)
SR
1.684*
[0.958]
-0.001
[0.249]
150
0.07
(4)
RP
1.501*
[0.833]
0.192
[0.234]
150
0.07
1.345
[1.197]
-0.035
[0.211]
0.71
[1.051]
0.142
[0.141]
0.711
[1.200]
-0.041
[0.125]
150
0.15
1.692**
[0.813]
-0.123
[0.080]
150
0.37
Note: R2R=Roads to Recovery, SF&C=Stronger Families and Communities, SR=Sustainable Regions,
RP=Regional Partnerships. Robust standard errors in brackets. ***, ** and * denote statistical significance
at the 1%, 5% and 10% levels respectively. Population density is square kilometres per person. Income is
median weekly family income ($000s).
4.2 Number of grants
I now turn to looking at the relationship between partisanship and the number of grants
delivered. The intuition for this approach is that, with quasi-rational voters, a politician
may gain more political capital from being able to announce a larger number of grants.
Alternatively, more grants may allow a politician to target a larger number of interest
groups within the electorate. Either scenario suggests the possibility that a politician may
gain more political advantage from announcing ten separate grants of $50,000 than a
16
single $500,000 grant. Since I do not have data on the number of Roads to Recovery
programs per electorate, I focus in this section only on the other three programs.
Panel A of Table 3 shows the number of grants per electorate, including a control for
demographic characteristics. Compared with non-government seats, National Party
electorates tend to have an additional 5.2 Stronger Families and Communities grants, and
an additional 3.4 Regional Partnerships grants. Liberal Party electorates have an
additional 4.3 Stronger Families and Communities grants, and an additional 1.7 Regional
Partnerships grants. There are no significant differences in the number of Sustainable
Regions grants allocated to either National Party or Liberal Party electorates.
In Panel B of Table 3, I add a control for the total amount of funding (in effect now
testing whether the funding in government electorates is delivered in smaller parcels than
in non-government electorates).13 For the most part, the results are qualitatively similar to
those in Panel A. Controlling for total funding, Liberal Party electorates receive a larger
number of Stronger Families and Communities grants and Regional Partnerships grants.
Controlling for total funding, National Party electorates receive a larger number of
Regional Partnerships grants, but a smaller number of Stronger Families and
Communities grants.
13
This specification is presented since it is easily comparable with the results in Panel A of Table 3. Results
are similar if the dependent variable is the average grant size.
17
Table 3: Partisanship and Number of grants
Dependent Variable: Number of grants per electorate
Panel A
(1)
(2)
SF&C
SR
National Party
5.195***
4.852
[1.907]
[3.946]
Liberal Party
4.338***
-0.214
[1.221]
[0.487]
Demographic
Yes
Yes
Controls
Observations
150
150
R-squared
0.25
0.16
Panel B
National Party
-1.631**
1.254
[0.760]
[0.853]
Liberal Party
0.713**
-0.119
[0.315]
[0.215]
Total funding ($m)
467.050***
2.675***
[18.401]
[0.307]
Demographic
Yes
Yes
Controls
Observations
150
150
R-squared
0.93
0.91
(3)
RP
3.355**
[1.311]
1.740***
[0.583]
Yes
150
0.46
2.465*
[1.482]
1.562***
[0.512]
1.255**
[0.608]
Yes
150
0.56
Note: SF&C=Stronger Families and Communities, SR=Sustainable Regions, RP=Regional Partnerships.
Robust standard errors in brackets. ***, ** and * denote statistical significance at the 1%, 5% and 10%
levels respectively. Demographic controls are a quadratic in median family income for SF&C, and a
quadratic in population density for SR and RP.
4.3 Swing Seats or Base?
An important question in the existing literature is whether targeted funding tends to be
delivered more towards swing seats or safe seats. To test this, I restrict the analysis to
Coalition seats, and regress the amount of funding on the Coalition’s share of the vote in
the 2001 election.14 The results are presented in Table 4. I find no evidence that any of
the programs are significantly more targeted towards marginal seats. Indeed, Roads to
14
Non-government seats are excluded from this analysis on the basis that it is not clear whether targeted
funding in non-government electorates would have led voters to reward the sitting member or the
governing party. Assuming that voters always reward the governing party for targeted funding, the
appropriate way to conduct the analysis would be to include all seats, and then regress the funding
allocation for each program (or the number of grants) on |Coalition Voteshare-50%|. Such an exercise
produces results that are qualitatively similar to those shown in Table 4.
18
Recovery funding appears to have been more generous towards safer seats. I obtain
similar results when the dependent variable is the number of grants instead of the total
funding allocation.
Note that while the results in Table 4 do not show any strong patterns according to
whether seats were safe or marginal, it would not be correct to say that funding was
spread evenly across Coalition seats. As the results in Tables 2 and 3 demonstrated, seats
held by the National Party, the junior partner in the Coalition, received significantly more
funding than those held by the Liberal Party.15
Table 4: Swing Seats or Base?
Sample is electorates won by the Coalition in 2001
Panel A: Dependent variable is total funding per electorate ($m)
(1)
(2)
(3)
R2R
SF&C
SR
Coalition Voteshare
0.596**
0.000
-0.038
[0.265]
[0.000]
[0.024]
Observations
85
85
85
R-squared
0.08
0
0.01
Panel B: Dependent variable is number of grants per electorate
Coalition Voteshare
0.031
-0.098
[0.178]
[0.069]
Observations
85
85
R-squared
0
0.01
(4)
RP
-0.005
[0.044]
85
0
0.151
[0.121]
85
0.02
Note: R2R=Roads to Recovery, SF&C=Stronger Families and Communities, SR=Sustainable Regions,
RP=Regional Partnerships. Robust standard errors in brackets. ***, ** and * denote statistical significance
at the 1%, 5% and 10% levels respectively.
5. An Empirical Analysis of the Effect of Funding on Elections
How do voters respond at the ballot box to targeted funding? To test this, I analyse the
relationship between additional funding provided through the programs analysed above
and the vote received by the Coalition in the 2004 election. However, as has been noted
by others, it is possible that the relationship between funding and voting is biased by
15
This result is consistent with the results of Tham and Grove (2004) for political donations. They found
that between 1999-2000 and 2001-02, the National Party received considerably more private donations per
vote ($28.64) than the Liberal Party ($18.62), Labor Party ($22.14), the Australian Democrats ($6.12) or
the Greens ($8.51).
19
omitted variables. The direction of this bias is in principle unclear. For example, the
quality of the Coalition candidate and targeted funding could be negatively correlated
(eg. if funding is centrally allocated, the government might devote more resources to
electorates where their candidate is of low quality). Alternatively, Coalition candidate
quality might be positively correlated with funding (eg. if the main determinant of
funding is the perseverance of the local Coalition MP).
To address this problem, I focus not on the share of the vote received by the Coalition
candidate in the 2004 election, but on the swing towards the Coalition. Since the
Coalition candidate in most electorates was the same in 2001 and 2004, the swing is more
likely to reflect factors that have changed between the two elections. Unobserved factors
(such as the quality of the Coalition candidate) are assumed to have affected the
Coalition’s voteshare equally in both elections.
Table 5 presents the results from regressing the swing towards the Coalition on the
amount of targeted funding. The coefficient on the Roads to Recovery program is 0.25,
indicating that a $1 million increase in funding raised the Coalition’s share of the vote by
0.25 percentage points. Since the standard deviation on Roads to Recovery funding is 9.7,
a one standard deviation increase in Roads to Recovery funding boosted the Coalition’s
voteshare by 2.4 percentage points. The coefficient on the Stronger Families and
Communities program is substantially larger, at 117.7. However, the standard deviation
of Stronger Families and Communities funding is much smaller. A one standard
deviation increase in Stronger Families and Communities funding ($0.016 million)
boosted the Coalition’s voteshare by 1.9 percentage points. The coefficients on the
Sustainable Regions and Regional Partnerships programs are not statistically significant.
One possibility is that the swing to the Coalition between 2001 and 2004 was driven by
factors that were also correlated with both funding allocations. To test this, Panel B
includes controls for population density (for Roads to Recovery, Sustainable Regions,
and Regional Partnerships), and for income (for Stronger Families and Communities).
This has the effect of increasing the size of the Roads to Recovery coefficient
20
substantially, and reducing the Stronger Families and Communities coefficient very
slightly. Both remain statistically significant at the 1 percent level.
Next, I include a control for the share of the vote received by the Coalition in 2001. As
the results in Table 4 showed, more Roads to Recovery funding was provided to safe
Coalition seats than marginal Coalition seats. It is also possibility that the Coalition’s
swing from the 2001 to 2004 election was positively correlated with its voteshare in 2001
(for example, government parliamentarians might have been more successful than
opposition representatives at exploiting the resources of incumbency). If this were the
case, then the relationship between funding and the swing might be spurious.
Such an approach could be over-cautious, as for example if one of the channels through
which Coalition incumbents were particularly successful was through obtaining targeted
funding. In Panel C, I include a control for the Coalition’s voteshare in 2001. This has the
effect of reducing the coefficients on Roads to Recovery and Stronger Families and
Communities to about one-third to one-quarter of their magnitude in Panel A. Both are
still significant at the 5 percent level. In Panel D, I include controls for both
demographics and voteshare. In this specification, Roads to Recovery expenditure is
positively correlated with the swing towards the Coalition. The other three programs are
no longer statistically significant.
21
Table 5: Electoral Swing and Funding Allocation
Dependent Variable: Swing to Coalition (percentage points)
Panel A: Funding only
(1)
(2)
(3)
(4)
R2R
SF&C
SR
RP
Funding ($m)
0.248***
117.686***
0.169
0.495
[0.040]
[31.640]
[0.136]
[0.350]
Observations
150
150
150
150
R-squared
0.18
0.10
0.00
0.02
Panel B: Controlling for relevant demographics
Funding ($m)
0.370***
113.009***
-0.051
0.09
[0.071]
[36.682]
[0.268]
[0.284]
Income
-26.282**
[11.274]
Income2
12.291***
[4.664]
Population Density
-2.884**
4.245***
4.020***
[1.452]
[1.443]
[1.473]
Population Density2
0.215
-0.418***
-0.401**
[0.146]
[0.153]
[0.156]
Observations
150
150
150
150
R-squared
0.21
0.13
0.07
0.07
Panel C: Controlling for past voteshare
Funding ($m)
0.058**
42.575**
-0.092
-0.261
[0.025]
[17.252]
[0.219]
[0.223]
Coalition vote (2001)
0.820***
0.798***
0.832***
0.849***
[0.059]
[0.070]
[0.068]
[0.061]
Observations
150
150
150
150
R-squared
0.71
0.67
0.66
0.66
Panel D: Controlling for demographics and past voteshare
Funding ($m)
0.132**
22.143
-0.114
-0.419
[0.055]
[20.555]
[0.232]
[0.273]
Income
3.401
[7.663]
Income2
-2.99
[3.150]
Population Density
-1.909
0.634
1.229
[1.698]
[1.093]
[1.091]
Population Density2
0.158
-0.066
-0.11
[0.160]
[0.113]
[0.109]
Coalition vote (2001)
0.809***
0.844***
0.821***
0.836***
[0.059]
[0.081]
[0.068]
[0.062]
Observations
150
150
150
150
R-squared
0.71
0.69
0.66
0.67
Note: R2R=Roads to Recovery, SF&C=Stronger Families and Communities, SR=Sustainable Regions,
RP=Regional Partnerships. Robust standard errors in brackets. ***, ** and * denote statistical significance
at the 1%, 5% and 10% levels respectively. Population density is square kilometres per person. Income is
median weekly family income ($000s). Coalition vote in 2001 is the election result in 2001, based on the
2004 boundaries (see section 3.5 for details).
Since allocations of funding under these four programs might be positively correlated
with one another, Table 6 presents results with all four programs included together. In the
22
first column, I find that without any other controls, the coefficient on the Roads to
Recovery program is positive (and significant at the 1 percent level), while Stronger
Families and Communities is positive (and significant at the 10 percent level). Regional
Partnerships funding is, surprisingly, negative (and significant at the 10 percent level).
Since the coefficient on the Regional Partnerships program was insignificant in Table 5,
it would be unwise to make much of this result.
Controlling for demographics and the Coalition’s share of the vote in 2001, it is notable
that the magnitude of the Roads to Recovery coefficient is very similar to that in Table 5.
By contrast, the coefficient on the Stronger Families and Communities program is
statistically insignificant in columns (2), (3) and (4). The coefficient on the Regional
Partnerships program is statistically insignificant with only demographic controls
(column 2), but negative and significant in columns (3) and (4).
23
Table 6: Electoral Swing and Funding Allocation
Dependent Variable: Swing to Coalition (percentage points)
R2R
SF&C
RP
SR
(1)
0.250***
[0.044]
51.937*
[30.930]
-0.453*
[0.243]
-0.16
[0.195]
Income
Income2
Population Density
Population Density2
(2)
0.343***
[0.073]
45.447
[34.409]
-0.309
[0.228]
-0.106
[0.183]
-19.869
[12.054]
10.070**
[4.929]
-2.310*
[1.310]
0.177
[0.125]
Coalition vote (2001)
Observations
R-squared
148
0.2
148
0.25
(3)
0.087***
[0.026]
33.85
[21.940]
-0.580**
[0.278]
-0.097
[0.165]
0.825***
[0.055]
148
0.73
(4)
0.090**
[0.045]
18.576
[23.447]
-0.650**
[0.324]
-0.084
[0.139]
3.504
[5.961]
-3.202
[2.481]
-0.944
[1.202]
0.096
[0.110]
0.890***
[0.061]
148
0.75
Note: R2R=Roads to Recovery, SF&C=Stronger Families and Communities, SR=Sustainable Regions,
RP=Regional Partnerships. Robust standard errors in brackets. ***, ** and * denote statistical significance
at the 1%, 5% and 10% levels respectively. Population density is square kilometres per person. Income is
median weekly family income ($000s). Coalition vote in 2001 is the election result in 2001, based on the
2004 boundaries (see section 3.5 for details).
Do electorates that receive a larger number of grants have a larger swing towards the
Coalition? In Table 7, I show the relationship between the swing towards the Coalition
and the number of grants, focusing on the three programs for which data are available. In
Panel A (with no other controls), I find that an extra Stronger Families and Communities
grant raised voteshare by 0.25 percentage points, while an additional Regional
Partnerships grant raised voteshare by 0.4 percentage points (both significant at the 1
percent level). These two results are still statistically significant when controlling for
either electorate demographics (Panel B) or the Coalition’s voteshare in 2001 (Panel C).
However, when controls for both electorate demographics and Coalition voteshare in
24
2001 are included (Panel D), only the number of Regional Partnerships grants remains
statistically significant (at the 10 percent level).
Note however that including a control for 2001 voteshare may be “overcontrolling”.
Assuming this to be the case, the results in Table 7 suggest that the number of grants does
matter at the ballot box. So far as I am aware, this paper is the first to document an effect
of the number of grants delivered on the vote. This suggests that future analyses of porkbarrelling should take into account not only the quantum of funding, but also the number
of grants that are delivered to each constituency.
25
Table 7: Electoral Swing and Number of Grants
Dependent Variable: Swing to Coalition (percentage points)
Panel A: Number of grants only
(1)
(2)
(3)
SF&C
SR
RP
Number of Grants
0.254***
0.066
0.399***
[0.071]
[0.048]
[0.100]
Observations
150
150
150
R-squared
0.11
0
0.12
Panel B: Controlling for demographics
Number of Grants
0.250***
-0.017
0.414***
[0.080]
[0.093]
[0.091]
Income
-26.883**
[10.943]
Income2
12.644***
[4.552]
Population Density
4.249***
1.617
[1.436]
[1.617]
Population Density2
-0.419***
-0.218
[0.151]
[0.161]
Observations
150
150
150
R-squared
0.15
0.07
0.15
Panel C: Controlling for past voteshare
Number of Grants
0.099**
-0.032
0.089**
[0.038]
[0.085]
[0.044]
Coalition vote (2001)
0.794***
0.833***
0.802***
[0.070]
[0.068]
[0.067]
Observations
150
150
150
R-squared
0.67
0.66
0.66
Panel D: Controlling for demographics and past voteshare
Number of Grants
0.058
-0.042
0.130*
[0.044]
[0.089]
[0.070]
Income
3.449
[7.506]
Income2
-2.942
[3.102]
Population Density
0.656
-0.188
[1.079]
[1.528]
Population Density2
-0.07
-0.006
[0.111]
[0.145]
Coalition vote (2001)
0.838***
0.821***
0.796***
[0.082]
[0.068]
[0.066]
Observations
150
150
150
R-squared
0.69
0.66
0.67
Note: R2R=Roads to Recovery, SF&C=Stronger Families and Communities, SR=Sustainable Regions,
RP=Regional Partnerships. Robust standard errors in brackets. ***, ** and * denote statistical significance
at the 1%, 5% and 10% levels respectively. Population density is square kilometres per person. Income is
median weekly family income ($000s). Coalition vote in 2001 is the election result in 2001, based on the
2004 boundaries (see section 3.5 for details).
26
6. Conclusion
This program analyses the distribution of four programs that fall into the broad category
of regional assistance programs. These types of programs account for a large (and in
some cases, growing) share of government budgets. Regional assistance programs
amount to over US$16 billion in the US, more than €30 billion in the European Union,
and over $4 billion in Australia. In allocating these resources, there is therefore
considerable potential for political considerations to take precedence over social and
economic factors.16
The distribution of funding under these programs appears to have been strongly skewed
towards electorates that were held by the Coalition government in 2001. Compared with
non-government seats, and controlling for relevant demographic characteristics, seats
held by the National Party received on average $6.8 million more under the Roads to
Recovery program and $15,000 more under Stronger Families and Communities
program. Liberal Party seats received $2.7 million more under the Roads to Recovery
program, and $8000 more under the Stronger Families and Communities program. As
well as receiving more money, the number of grants was also higher in Coalition seats.
National Party electorates received an additional 5.2 Stronger Families and Communities
grants, and an additional 3.4 Regional Partnerships grants. Liberal Party electorates
received an additional 4.3 Stronger Families and Communities grants, and an additional
1.7 Regional Partnerships grants.
I find no evidence that the grants were more generous to more marginal electorates
(indeed, the Roads to Recovery program seems to have gone more to safe seats). To the
extent that funding was targeted among Coalition seats, it appears to have been towards
seats held by the smaller party in the Coalition, the National Party, and not towards swing
seats. This is surprising, in light of the fact that earlier studies of pork-barrelling in
16
US and EU figures are taken from Milligan and Smart (2005). Australia figure is the sum of expenditure
on agriculture, forestry, fisheries, transport and communication. While this may include some expenditure
that is not regionally directed, it also does not include expenditure in other portfolio areas (eg. recreation
and culture, tourism promotion, community assistance) that is regionally directed.
27
Australia have found the funding to be more generously targeted towards marginal
electorates (Gaunt 1999 and Denemark 2000 for sports grants in the early-1990s;
Andrews, Fry and Jakee 2005 for unemployment programs in the late-1990s).
Although this study is not the first to observe targeted assistance in a parliamentary
system which is not targeted at swing seats (Milligan and Smart 2005 saw a similar
pattern for Canadian regional development grants), such a finding differs from the usual
predictions of a model in which pork-barrelling in parliamentary systems tends to focus
on swing seats. One possible explanation is that during the particular 2001-2004 election
cycle, the government was always confident of winning re-election. Leigh and Wolfers
(2006) show that the betting market – an accurate prediction of election outcomes – had
the government as favourites from when they opened in July 2003 until polling day in
October 2004. Similarly, macroeconomic models of election forecasting (based on
variables such as unemployment and inflation), suggested that the Coalition would win
comfortably. Another possibility is that the government was more concerned about
assuaging internal tensions among the Coalition parties by providing largesse to National
Party electorates than about increasing its parliamentary majority.
Analysing the effect of the programs on the swing towards the Coalition in the 2004
election, I find robust evidence that additional funding increased the swing towards the
Coalition government, and suggestive evidence that a larger number of grants delivered
to an electorate also helped the government. In terms of the amount of funding delivered,
the most robust results are for the Roads to Recovery program. For every additional $1
million in Roads to Recovery funding, the Coalition increased its voteshare by between
0.06 and 0.37 percentage points, depending on the controls included in the regression.
Since the average number of votes cast per electorate in 2004 was 82,367, this suggests
that each additional vote obtained through the Roads to Recovery program cost between
$20,234 and $3281.17 This figure is similar to Levitt and Snyder (1997), whose estimates
17
In an opinion piece, I reported this figure as $28,000 (Leigh 2005). This regression differed from those
reported here in two respects: it excluded Roads to Recovery funding to the territories; and it was based on
swing data comparing the actual 2001 result with the 2004 result. Here, I include funding allocations to the
28
of the relationship between federal funding and voting suggest that each additional US
vote costs around US$14,000 (approximately $22,000 in 2004 Australian dollars).
Economists and political scientists in parliamentary democracies have traditionally been
less concerned about the issue of pork-barrelling than their US counterparts. These results
suggest that outside the US, more attention should be given to considering the role that
partisanship plays in funding decisions, and the role of targeted funding in shaping
election outcomes. Pork-barrelling is both inefficient and inequitable, and may have other
social costs, such as reducing the level of turnover among politicians below the social
optimum.18
four territory seats, and calculate the swing by comparing the 2001 result (translated onto 2004 electoral
boundaries) with the 2004 result.
18
In the 2004 Australian election, the incumbent re-election rate was 92% (130/141 incumbents seeking reelection were successful: Jackman 2005), which was lower than in the 2004 US congressional elections,
where the incumbent re-election rate was 99% (399/401). In the 2001 Australian federal election, the reelection rate was 96% (130/135), while in the 1998 election, it was 85% (109/129).
29
References
Alvarez, R. Michael and Jason L. Saving. 1997a. “Congressional Committees and the
Political Economy of Federal Outlays” Public Choice 92:55-73
Alvarez, R. Michael and Jason L. Saving. 1997b. “Deficits, Democrats, and Distributive
Benefits” Political Research Quarterly 50(4): 809-831
Andrews, Luke, Tim R.L. Fry, and Keith Jakee. 2005. “Compulsory Voting: Some
Analytics and a Test from Australia” mimeo, Royal Melbourne Institute of Technology:
Melbourne
Cadot, O., L-H Röller and A. Stephan. 2002. “Contribution to productivity or pork
barrel? The two faces of infrastructure investment,” Discussion Paper FS IV 02-09,
Wissenschaftszentrum Berlin.
Cain, Bruce, John Ferejohn, and Morris Fiorina. 1984. “The Constituency Service Basis
of the Personal Vote for U.S. Representatives and British Members of Parliament.”
American Political Science Review 78(1): 110-125.
Cox, Gary W. and Mathew D. McCubbins. 1986. “Electoral Politics as a Redistributive
Game.” Journal of Politics 48(2): 370-389.
Crampton, E. 2004. “Distributive politics in a strong party system: Evidence from
Canadian job grant programs,” Discussion paper, University of Canterbury.
Dahlberg, M and E. Johansson. 2002. “On the Vote-Purchasing Behavior of Incumbent
Governments,” American Political Science Review, 96, 27–40.
Dasgupta, S., A. Dhillon and B. Dutta. 2002. “Electoral goals and center-state transfers:
A theoretical model and empirical evidence from India,” Discussion paper, University of
Warwick.
Davis, Mark. 2005. “Recovery scheme is a one way street”. Australian Financial Review.
7 March
Denemark, David. 2000. “Partisan Pork Barrel in Parliamentary Systems: Australian
Constituency-Level Grants” Journal of Politics 62(3): 896-915
Department of Family and Community Services (FACS). 2005a. “Stronger Families and
Communities Strategy - Local Answers and Volunteer Small Equipment Grants $151
Million Over Five Years” FaCS 2047.0507. FACS: Canberra, ACT
Department of Family and Community Services (FACS). 2005b. “Stronger Families and
Communities Strategy - National Evaluation of the Stronger Families and
Communities Strategy”. FaCS 2049.0507. FACS: Canberra, ACT
30
Department of Transport and Regional Services (DOTARS). 2001a. National Black Spot
Programme Notes on Administration. DOTARS: Canberra, ACT
Department of Transport and Regional Services (DOTARS). 2001b. Stronger Regions, A
Stronger Australia. DOTARS: Canberra, ACT
Department of Transport and Regional Services (DOTARS). 2004. Roads to Recovery
Programme Annual Report 2003–2004. DOTARS: Canberra, ACT
Department of Transport and Regional Services and the Australian Local Government
Association (DOTARS/ALGA). 2003.Report on the Roads to Recovery Programme.
DOTARS: Canberra, ACT
Dixit, Avinash and John Londregan. 1996. “The Determinants of Success of Special
Interests in Redistributive Politics.” Journal of Politics 58(4): 1132-1155.
Feldman, Paul and James Jondrow. 1984. “Congressional Elections and Local Federal
Spending” American Journal of Political Science 28: 147-164
Flake, Jeff. 2006. “Earmarked Men” New York Times, 9 February
Gaunt, Clive. 1999. “Sports Grants and the Political Pork Barrel: An Investigation of
Political Bias in the Administration of Australian Sports Grants” Australian Journal of
Political Science 34(1): 63-74
Gordon, John S. 1993. “USS Boondoggle” American Heritage Magazine 44(1)
Jackman, Simon. 2005. “Incumbency Advantage and Candidate Quality” in Mortgage
Nation: The 2004 Australian Election. Marian Simms and John Warhurst (eds). 2005.
Perth, Western Australia: API Network/Curtin University of Technology. 335-347
Kopras, Andrew. 2003. Electorate Rankings: Census 2001 (Research Paper No.2
2002-03).
Parliamentary
Library:
Canberra,
ACT.
Available
at
http://www.aph.gov.au/library/pubs/rp/2002-03/03RP02.htm
Lancaster, Thomas, and W. David Patterson. 1990. “Comparative Pork Barrel Politics:
Perceptions from the West German Bundestag.” Comparative Political Studies 22(4):
458-477.
Leigh, Andrew. 2005. “Program paved the road to re-election”, Australian Financial
Review, March 10, 71
Leigh, Andrew and Wolfers, Justin. 2006. “Competing Approaches to Forecasting
Elections: Economic Models, Opinion Polling and Prediction Markets” Economic
Record, forthcoming
31
Levitt, Steven D. and James M. Snyder. 1997. “The Impact of Federal Spending on
House Election Outcomes.” Journal of Political Economy 105(1): 30–53.
Lindstädt, René. 2005. “Core Supporters or Swing Voters? The Geo-Politics of New Deal
Spending” mimeo, Washington University: St Louis, MO
Milligan, Kevin and Michael Smart. 2005. “Regional Grants as Pork Barrel Politics”.
CESifo Working Paper 1453
Senate Finance and Public Administration Committee (SFPAC). 2005. Regional
Partnerships and Sustainable Regions Programs. SFPAC: Canberra, ACT
Stein, Robert M., and Kenneth N. Bickers. 1994. “Congressional Elections and the Pork
Barrel” Journal of Politics 56(2): 377-399
Stein, Robert M. and Kenneth N. Bickers. 1995. Perpetuating the Pork Barrel Cambridge
University Press: Cambridge, UK
Tham, Joo-Cheong and Grove, David. 2004. “Public Funding and Expenditure
Regulation of Australian Political Parties: Some Reflections” Federal Law Review 32(3):
397-424
32
Data Appendix
Electorate
Adelaide (SA)
Aston (Vic)
Ballarat (Vic)
Banks (NSW)
Barker (SA)
Barton (NSW)
Bass (Tas)
Batman (Vic)
Bendigo (Vic)
Bennelong (NSW)
Berowra (NSW)
Blair (Qld)
Blaxland (NSW)
Bonner (Qld)
Bonython (SA)
Boothby (SA)
Bowman (Qld)
Braddon (Tas)
Bradfield (NSW)
Brand (WA)
Brisbane (Qld)
Bruce (Vic)
Burke (Vic)
Calare (NSW)
Calwell (Vic)
Canberra (ACT)
Canning (WA)
Capricornia (Qld)
Casey (Vic)
Party
(2001)
LIB
LIB
ALP
ALP
LIB
ALP
ALP
ALP
ALP
LIB
LIB
LIB
ALP
ALP
ALP
LIB
LIB
ALP
LIB
ALP
ALP
ALP
ALP
IND
ALP
ALP
LIB
ALP
LIB
Coalition
Voteshare
(2001)
50.22
56.17
47.27
47.11
65.69
43.98
47.94
24.92
46.43
57.71
65.65
58.5
34.79
49.05
39.58
57.35
48.58
44.04
71.16
39.95
46.87
44.45
44.49
51.73
32.27
40.56
50.38
43.14
57.16
Swing
(%)
-1.63
10.15
-0.63
0.39
11.58
-4.54
3.68
-8.77
0.84
0.48
4.36
7.91
-5.27
1.46
6.65
1.67
7.57
-9.47
21.51
0.35
-3.40
-0.23
-0.25
-0.63
-0.59
-4.92
9.34
-2.39
7.75
33
Persons
per km2
1769.8
1279.8
13.7
2245.7
2.4
3207.3
12.4
2336.3
12.1
2412
187.1
7.7
2427.3
632.25
531.4
1007.5
212.1
7.8
1321.1
276.1
2074.1
1877.8
41.5
5.8
482.2
80.5
27.8
0.5
298.7
Mean
weekly
inc ($)
1011
1180
793
1065
730
1067
760
864
736
1300
1488
765
839
1035.5
718
1003
972
688
1759
752
1208
1003
993
873
964
1347
915
886
1058
R2R
Funding
($m)
4.33
2.37
11.54
1.76
17.47
1.94
7.76
2.18
13.38
2.09
2.22
11.13
2.07
3.96
4.12
4.12
8.88
2.32
4.16
3.73
2.47
14.77
3.09
10.00
6.64
16.19
4.70
SF&C
Funding
($m)
0.02
0.01
0.02
0.00
0.01
0.00
0.02
0.01
0.03
0.00
0.01
0.02
0.00
SF&C
Projects
(#)
7
9
12
1
9
0
9
4
17
2
5
9
0
SR
Funding
($m)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.12
0.00
SR
Projects
(#)
0
0
0
0
0
0
0
0
0
0
0
1
0
RP
Funding
($m)
0.02
0.01
1.03
0.00
1.23
0.00
0.71
0.00
0.52
0.00
0.00
1.64
0.00
RP
Programs
(#)
1
1
7
0
16
0
5
0
5
0
0
7
0
0.01
0.01
0.00
0.01
0.01
0.00
0.02
0.01
0.03
0.02
0.01
0.03
0.01
0.03
0.00
7
11
2
3
4
0
12
3
15
9
3
12
4
19
5
0.00
0.00
0.00
0.18
0.18
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0
0
0
5
5
0
0
0
0
0
0
0
0
0
0
0.03
0.00
0.03
0.07
0.07
1.23
0.34
0.00
0.08
1.95
0.00
0.00
3.60
4.33
0.02
1
0
1
1
1
6
2
0
2
5
0
0
5
10
1
Electorate
Charlton (NSW)
Chifley (NSW)
Chisholm (Vic)
Cook (NSW)
Corangamite (Vic)
Corio (Vic)
Cowan (WA)
Cowper (NSW)
Cunningham (NSW)
Curtin (WA)
Dawson (Qld)
Deakin (Vic)
Denison (Tas)
Dickson (Qld)
Dobell (NSW)
Dunkley (Vic)
Eden-Monaro (NSW)
Fadden (Qld)
Fairfax (Qld)
Farrer (NSW)
Fisher (Qld)
Flinders (Vic)
Forde (Qld)
Forrest (WA)
Fowler (NSW)
Franklin (Tas)
Fraser (ACT)
Fremantle (WA)
Gellibrand (Vic)
Gilmore (NSW)
Party
(2001)
ALP
ALP
ALP
LIB
LIB
ALP
ALP
NP
GRN
LIB
NP
LIB
ALP
LIB
LIB
LIB
LIB
LIB
LIB
LIB
LIB
LIB
LIB
LIB
ALP
ALP
ALP
ALP
ALP
LIB
Coalition
Voteshare
(2001)
43.34
34.71
47.23
64
55.67
41.3
44.49
54.73
39.35
63.91
57.99
51.74
35.74
55.97
50.38
55.42
51.69
62.29
59.21
66.37
62.06
57.62
57.38
57.61
28.51
41.96
37.31
39.33
28.22
64.63
Swing
(%)
-4.57
-5.33
-1.30
6.82
2.62
-1.39
1.97
4.10
-0.81
7.67
6.38
4.17
-6.14
4.83
5.70
6.78
1.29
8.73
6.23
11.62
7.08
7.41
9.48
6.65
-10.61
-3.59
-6.98
-2.41
-4.75
2.78
34
Persons
per km2
206.7
1199.4
2102.7
712.1
15.3
150.5
614.8
15.6
291.3
1380.8
6.4
1947
420.3
179.5
137.2
953.9
4.5
220.4
33.2
1.3
231.9
61.9
29.2
5.7
2855.7
14.3
350.8
588.9
1690.6
20.8
Mean
weekly
inc ($)
848
899
1065
1296
894
825
992
618
1003
1381
830
1057
861
1071
877
948
792
926
677
816
700
828
847
845
714
810
1304
937
860
698
R2R
Funding
($m)
2.88
2.81
2.29
1.90
10.94
5.51
3.92
7.96
2.38
3.35
9.09
2.26
3.35
4.08
3.14
4.42
12.05
3.76
4.27
29.85
4.81
8.56
6.12
15.40
2.45
4.53
10.00
3.33
2.24
6.43
SF&C
Funding
($m)
0.01
0.01
0.00
0.01
0.04
0.04
0.00
0.02
0.02
0.01
0.02
0.01
0.00
0.01
0.01
0.01
0.10
0.01
0.01
0.04
0.01
0.04
0.04
0.05
0.01
0.01
0.02
0.01
0.01
0.04
SF&C
Projects
(#)
3
2
2
4
25
21
1
8
9
2
11
7
1
4
6
5
49
2
3
26
6
17
17
23
4
8
7
4
5
16
SR
Funding
($m)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
SR
Projects
(#)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
RP
Funding
($m)
0.00
0.03
0.00
0.00
0.33
0.17
0.06
0.88
0.25
0.20
1.12
0.00
0.13
0.00
1.64
0.30
1.50
0.21
0.92
1.61
0.00
0.02
0.93
2.57
0.00
0.00
0.00
0.19
0.08
0.68
RP
Programs
(#)
0
1
0
0
1
3
1
9
2
1
9
0
2
0
4
2
12
1
4
12
0
1
4
16
0
0
0
3
1
11
Electorate
Gippsland (Vic)
Goldstein (Vic)
Gorton (Vic)
Grayndler (NSW)
Greenway (NSW)
Grey (SA)
Griffith (Qld)
Groom (Qld)
Gwydir (NSW)
Hasluck (WA)
Herbert (Qld)
Higgins (Vic)
Hindmarsh (SA)
Hinkler (Qld)
Holt (Vic)
Hotham (Vic)
Hughes (NSW)
Hume (NSW)
Hunter (NSW)
Indi (Vic)
Isaacs (Vic)
Jagajaga (Vic)
Kingsford Smith (NSW)
Kalgoorlie (WA)
Kennedy (Qld)
Kingston (SA)
Kooyong (Vic)
La Trobe (Vic)
Lalor (Vic)
Leichhardt (Qld)
Party
(2001)
NP
LIB
ALP
ALP
ALP
LIB
ALP
LIB
NP
ALP
LIB
LIB
LIB
NP
ALP
ALP
LIB
LIB
ALP
LIB
ALP
ALP
ALP
LIB
IND
ALP
LIB
LIB
ALP
LIB
Coalition
Voteshare
(2001)
58.05
59.48
40
28.71
46.89
60.56
44.34
65.09
64.88
48.22
51.62
58.39
51.86
50.04
36.68
38.98
60.41
59.79
39.14
61.15
47.19
44.36
41.1
54.34
58.95
47.58
60.94
53.67
34.37
56.39
Swing
(%)
6.40
5.28
-4.90
-11.95
2.13
8.52
-5.53
11.40
10.94
2.72
5.45
4.56
-0.56
3.71
2.44
-1.90
5.84
9.23
-8.30
10.94
1.82
-1.75
-4.56
4.15
0.00
0.72
4.35
3.98
-2.74
6.80
35
Persons
per km2
3.2
2506.8
780.4
4465.7
1145.8
0.1
1052.4
19.3
0.7
524.2
48.3
3183
1790.3
8
1511
1727.2
470.1
4.8
11.6
5.1
613.8
1565.7
2576.5
0.1
0.3
729.5
2485.3
249.3
222
1.1
Mean
weekly
inc ($)
700
1385
950
1190
1107
705
1099
838
716
898
960
1570
866
724
896
951
1397
927
874
809
981
1168
1146
1072
790
825
1593
1089
988
883
R2R
Funding
($m)
17.05
2.15
2.62
1.88
2.57
22.02
4.04
10.16
41.03
4.30
4.53
3.40
4.07
13.20
1.69
2.24
2.17
14.14
7.41
18.44
2.50
2.96
1.85
48.87
28.68
5.00
2.06
4.27
3.65
9.25
SF&C
Funding
($m)
0.05
0.00
SF&C
Projects
(#)
26
3
SR
Funding
($m)
12.93
0.00
SR
Projects
(#)
44
0
RP
Funding
($m)
0.59
0.00
RP
Programs
(#)
3
0
0.00
0.01
0.04
0.01
0.03
0.08
0.00
0.02
0.00
0.01
0.04
0.00
0.01
0.00
0.03
0.02
0.05
0.01
0.01
0.00
0.04
0.02
0.01
0.00
0.03
0.00
0.03
2
5
17
6
18
30
1
6
2
4
11
1
3
1
14
10
32
6
5
1
12
9
5
2
17
1
13
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.42
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.81
10.48
0.00
0.00
0.00
0.00
0.58
0
0
0
0
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
0
19
31
0
0
0
0
3
0.00
0.00
4.25
0.00
0.07
4.15
0.09
0.20
0.00
0.11
0.23
0.00
0.00
0.01
0.53
0.26
1.98
0.00
0.09
0.10
5.93
7.44
0.34
0.03
0.21
0.28
0.90
0
0
13
0
1
24
1
3
0
1
4
0
0
1
4
1
14
0
1
1
33
10
3
1
2
2
6
Electorate
Lilley (Qld)
Lindsay (NSW)
Lingiari (NT)
Longman (Qld)
Lowe (NSW)
Lyne (NSW)
Lyons (Tas)
Melbourne Ports (Vic)
Macarthur (NSW)
Mackellar (NSW)
Macquarie (NSW)
Makin (SA)
Mallee (Vic)
Maranoa (Qld)
Maribyrnong (Vic)
Mayo (SA)
McEwen (Vic)
McMillan (Vic)
McPherson (Qld)
Melbourne (Vic)
Menzies (Vic)
Mitchell (NSW)
Moncrieff (Qld)
Moore (WA)
Moreton (Qld)
Murray (Vic)
New England (NSW)
Newcastle (NSW)
North Sydney (NSW)
O'Connor (WA)
Party
(2001)
ALP
LIB
ALP
LIB
ALP
NP
ALP
ALP
LIB
LIB
LIB
LIB
NP
NP
ALP
LIB
LIB
LIB
LIB
ALP
LIB
LIB
LIB
LIB
LIB
LIB
IND
ALP
LIB
LIB
Coalition
Voteshare
(2001)
45.17
55.47
44.71
52.72
46.19
61.24
41.83
44.31
56.96
66.87
58.67
53.76
69.93
66.01
32.62
62.87
51.2
47.54
62.55
29.91
58.94
71.32
65.42
56.04
54.21
73.93
63.85
43.09
63.22
69.09
Swing
(%)
-2.97
2.51
-5.01
6.41
-1.40
7.43
0.42
-0.89
6.01
7.30
4.57
-0.92
14.30
13.23
-1.77
6.44
5.27
3.59
7.79
-11.19
6.22
10.03
11.99
7.83
2.87
13.13
-0.64
-6.53
3.43
10.84
36
Persons
per km2
884.9
358.8
0.1
64.6
3150.7
13.4
2.1
3038.5
223.3
575.4
30.5
1137.2
1.7
0.2
1817.5
59.7
9.6
20.2
358.7
3040.8
969.6
683.5
915.7
1427.9
1193.9
7.6
2.3
733.8
3476
0.7
Mean
weekly
inc ($)
958
1132
856
720
1270
642
685
1406
1088
1416
1077
930
755
797
893
1000
945
817
789
1154
1241
1597
845
1131
1011
813
741
836
1792
723
R2R
Funding
($m)
4.17
2.78
10.00
4.36
1.98
9.94
15.07
1.88
2.70
2.21
5.11
3.96
25.26
59.64
2.31
7.24
11.07
14.00
3.28
2.43
2.58
2.52
4.51
2.62
4.10
19.43
23.01
2.70
2.08
56.42
SF&C
Funding
($m)
0.01
0.00
0.01
0.04
0.01
0.02
0.02
0.01
0.01
0.02
0.03
0.00
0.03
0.03
0.01
0.01
0.02
0.02
0.01
0.02
0.01
0.00
0.01
0.00
0.00
0.04
0.01
0.00
0.00
0.04
SF&C
Projects
(#)
7
2
5
19
6
13
8
4
3
5
17
1
16
15
5
3
13
14
3
11
3
1
4
3
1
21
7
0
0
17
SR
Funding
($m)
0.00
0.00
0.00
0.00
0.00
0.00
0.60
0.00
7.95
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
SR
Projects
(#)
0
0
0
0
0
0
3
0
12
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
RP
Funding
($m)
0.06
0.07
5.47
0.73
0.00
3.21
0.50
0.00
0.00
0.00
0.08
0.19
0.45
1.80
0.00
0.38
1.40
0.12
0.46
0.33
0.00
0.00
0.06
0.00
0.12
1.51
5.79
0.10
0.00
7.12
RP
Programs
(#)
1
1
8
2
0
9
7
0
0
0
1
1
11
12
0
3
10
2
4
3
0
0
1
0
2
8
11
3
0
22
Electorate
Oxley (Qld)
Page (NSW)
Parkes (NSW)
Parramatta (NSW)
Paterson (NSW)
Pearce (WA)
Perth (WA)
Petrie (Qld)
Port Adelaide (SA)
Prospect (NSW)
Rankin (Qld)
Reid (NSW)
Richmond (NSW)
Riverina (NSW)
Robertson (NSW)
Ryan (Qld)
Scullin (Vic)
Shortland (NSW)
Solomon (NT)
Stirling (WA)
Sturt (SA)
Swan (WA)
Sydney (NSW)
Tangney (WA)
Throsby (NSW)
Wakefield (SA)
Wannon (Vic)
Warringah (NSW)
Watson (NSW)
Wentworth (NSW)
Party
(2001)
ALP
NP
NP
LIB
LIB
LIB
ALP
LIB
ALP
ALP
ALP
ALP
NP
NP
LIB
LIB
ALP
ALP
CLP
ALP
LIB
ALP
ALP
LIB
ALP
ALP
LIB
LIB
ALP
LIB
Coalition
Voteshare
(2001)
41.86
52.77
58.74
51.15
51.42
56.88
38.79
53.42
35.35
37.19
43.32
33.13
51.68
69.87
56.98
58.62
30.81
41.22
50.09
48.42
58.18
47.96
34.96
57.97
34.9
64.56
59.61
62.66
32.69
57.86
Swing
(%)
-5.72
2.83
10.05
-1.37
6.27
9.49
-1.13
6.17
-4.91
-0.72
-2.03
-4.32
-1.04
10.71
3.31
5.62
-4.69
-5.09
2.76
2.84
2.55
0.92
-8.92
7.75
-7.45
1.32
7.77
4.13
-6.49
1.53
37
Persons
per km2
201.5
7.3
0.5
2351.4
12.3
4.6
1701.1
853
500.5
914.6
948
2103.7
57.2
2.9
168
530.9
1489.3
625
315.2
1573.1
1755.1
1110.2
3694.8
1721.5
283.1
4.8
3.8
2297
3907.7
4971
Mean
weekly
inc ($)
839
656
757
1080
711
885
909
912
760
1019
853
827
654
866
908
1305
970
767
1182
888
983
890
1517
1127
859
774
780
1572
887
1649
R2R
Funding
($m)
5.16
12.26
34.91
2.47
7.51
15.38
4.19
3.79
4.48
2.52
4.82
2.20
6.72
23.24
3.04
4.02
2.75
3.01
10.00
2.95
4.02
3.44
2.01
3.35
2.47
7.61
26.80
2.01
1.89
1.72
SF&C
Funding
($m)
0.01
0.03
0.03
0.01
0.05
0.03
0.02
0.00
0.00
0.01
0.01
0.00
0.02
0.02
0.02
0.00
0.01
0.01
0.01
0.01
0.01
0.02
0.01
0.00
0.00
0.02
0.03
0.00
0.00
0.01
SF&C
Projects
(#)
5
13
14
6
23
13
10
0
2
2
2
1
9
9
8
1
4
4
4
2
4
7
5
1
1
10
20
1
2
7
SR
Funding
($m)
0.00
3.77
0.00
0.00
0.00
0.00
0.00
0.00
0.43
0.00
0.00
0.00
2.92
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
3.90
0.00
0.00
0.00
0.00
SR
Projects
(#)
0
12
0
0
0
0
0
0
1
0
0
0
16
0
0
0
0
0
0
0
0
0
0
0
0
7
0
0
0
0
RP
Funding
($m)
0.08
0.43
1.39
0.00
0.75
1.67
0.20
0.43
0.04
0.00
0.00
0.00
12.14
0.41
0.41
0.00
0.10
0.00
0.80
0.00
0.00
0.11
0.13
0.00
0.39
0.66
0.86
0.00
0.00
0.22
RP
Programs
(#)
1
8
6
0
7
3
2
4
1
0
0
0
6
4
5
0
1
0
3
0
0
1
2
0
2
8
6
0
0
1
Coalition
Mean
R2R
SF&C
SF&C
SR
SR
RP
Swing
Persons
Voteshare
weekly
Funding Funding Projects Funding Projects Funding
2
(%)
per km
(2001)
inc ($)
($m)
($m)
(#)
($m)
(#)
($m)
Werriwa (NSW)
ALP
41.51
-5.06
839
1031
2.60
0.00
2
2.56
2
0.12
Wide Bay (Qld)
NP
60.73
7.94
2.3
622
10.01
0.04
17
5.33
14
0.51
Wills (Vic)
ALP
30.58
-6.60
2421.9
917
2.36
0.01
6
0.00
0
0.17
Note: R2R=Roads to Recovery, SF&C=Stronger Families and Communities, SR=Sustainable Regions, RP=Regional Partnerships.
ALP=Australian Labor Party, CLP=Country Liberal Party, GRN=Green Party, IND=Independent, LIB=Liberal Party of Australia, NP=National Party of
Australia.
Electorate
Party
(2001)
38
RP
Programs
(#)
1
2
1